Cyber Security for Industrial Control System using Deep Neural Network Model

Machine learning, Deep Learning and cyber security have an enormous multidisciplinary convergence. Recent advances in deep learning have indicated a tremendous increase in cyber security and intrusion detection. Through the development of concepts, cyber security specialists are enhancing the safety of Deep Learning models for the cyber environment. Before the development of deep learning and machine learning, cybercriminals conducted their attacks using Petri net formalizations. However, as technology advanced, they began to employ more sophisticated methods. The focus of this paper is on deep learning CNN models for industrial cyber control systems using network traffic data. Deep transfer learning is applied by including additional characteristics either from the input data or through pre-processing methods. In addition, supervised learning is done to design a self-updating model based on the various attack variations and suspicious behaviour is detected. Finally, it is further illustrated that deep learning is being used in a variety of fields beyond the cyber security, including education, robotics, automation, and many more.